Hi, I'm Howard Marks, and this is How to Think About Risk. The title of this class is How to Think About Risk. That's an important title. Not what to think, how to think. The first question is, what is risk?
Risk, in my opinion, is the ultimate test of an investor's skill. The return... alone doesn't tell you how good a job the manager did. The key question is, you see the return, you must ask, how much risk did the manager bear to get that return? Now let's look at this range of managers.
What we posit is that the market is up 10% or down 10%. Now, let's look at some individual managers. The first one is up 10 when the market's up 10, and down 10 when the market's down 10. Accomplishes nothing.
You might as well have invested in an index fund that emulates the performance of the market. No skill, no value added. Now, let's look at the second one.
Up 20 when the market goes up 10, down 20 when the market goes down 10. No skill, no value added, just a lot of aggressiveness. What about this one? Up 5 when the market's up 10, down 5 when the market's down 10. Again, no selection ability, no discernment, no value added, just defensiveness.
You don't need help in achieving that, and you sure shouldn't pay a lot for it. But what about the next one? He or she is up 15 when the market's up 10 and down 10 when the market's down 10. So in other words, market type losses on the downside. But superior gains on the upside.
Value-added, what I call asymmetry, does better in the good times than does poorly in the bad times. But what about this one? And I think this is the most interesting.
I think maybe it characterizes me. Maybe it characterizes Oaktree to some extent. Up 10 when the market's up 10, down 5 when the market's down 10. So market-type gains in the good times, I personally think that performing— with the market when it does well is good enough. And that's almost all the time.
Nobody should have to beat the market when it does well. But if you can do that and at the same time be ready to decline less when the market has its down spells, I think that's accomplishing something very important. I believe that risk is not volatility. The academics developing investment theory, largely at the University of Chicago in the early 60s, just a couple of years before I got there, adopted volatility as their measure of risk.
I believe they did so largely because because volatility is readily quantifiable and nothing else is. I think that volatility can be an indicator of the presence of risk, a symptom if you will, but it's not risk itself. So if risk is not volatility, then what is it?
And in my opinion, and in the real world sense, risk is the probability of loss. I think this is what most people mean when they say risk, and I think that this is what people demand compensation for if they're going to bear it. Nobody sitting around at Oaktree says, well, you know, we shouldn't make that investment because it might be volatile, or because it might be volatile, we should demand a higher return. No, they say that about the possibility of loss.
We're not going to make that investment because the possibility of loss is too high or because of the possibility of loss we're going to demand a risk premium in terms of the return. This is risk, the possibility of loss. Now another important question is, is risk quantifiable in advance?
And I believe it's not. Like most things occurring in the future, risk cannot be anything except a matter of opinion. I was writing my first memo about risk in 2006. I wrote about my belief that risk is not quantifiable in advance. And then I hit the return key and went on to the next section and wrote something I had never thought about before. My belief that risk is unquantifiable even after the fact.
And I think this is a fascinating topic. You buy something for a dollar and a year later you sell it for two dollars. Because it risks. And the interesting thing is that you can't tell from the outcome.
A profitable investment may or may not have been risky. Was it a safe investment that in the case of my example was sure to double, or was it a risky investment where you got lucky in terms of the outcome? And as I say, you can't tell from the outcome. The bottom line is to me it's impossible to quantify risk in advance. advance or even in hindsight.
The possibility of loss is not the only form of risk. There are lots of forms of risk. My last memo on the general subject, it's called Risk Revisited Again, and I talk in there about 24 or 25 different forms of risk, some serious, some...
facetious, some important, some less important and obscure, but still it comes in many forms. The risk of missing opportunities is another important risk. In other words, if you think about it, the risk of not taking enough risk.
Another really important form of risk. I think one of the key risks in investing is the chance of being forced out at the bottom, which is a big mistake. Buying at the high and seeing a decline or selling out at the low and missing out on the recovery.
Clearly, it's the latter. If you buy at a high and you experience a decline, if you're able to hold throughout and not lose your nerve, the next high is usually. higher than the last high. The fact that you experienced a downward fluctuation might have been uncomfortable for a little while, but by the time the new high is achieved, you're back to your cost and more.
But if you sell at the bottom and miss out on the subsequent recovery, that means you've gotten off the track of investing and may never get back on, in my opinion, selling at the bottom. It's the cardinal sin in investing. Now I want to get a little philosophical. One of my great heroes, Peter Bernstein, probably the best thinker in a philosophic sense and a real investment sage who sadly passed away around 2009, once said, Essentially, risk says we don't know what's going to happen. We walk every moment into the unknown.
He said there's a range of outcomes and we don't know where the actual outcome is going to fall within the range. And often, we don't know what the range is. So in other words, we have ignorance to varying degrees about what the future holds. And it is from this ignorance that risk ensues.
If we knew what was going to happen, by definition, there would be no risk. In the memo that I mentioned before, Risk Revisited, again, there's a great quote that I got from a Peter Bernstein memo. And I thought it was so important that I took it over word for word.
in my memo. It's from G.K. Chesterton, who was an English writer, and he said the following, and I'm going to give it to you word for word because it's so important. The real trouble with this world of ours is not that it is an unreasonable world or even that it is a reasonable one. The commonest kind of trouble is that it is nearly reasonable but not quite.
Life is not an illogicality, yet it is a trap for logicians. It looks just a little more mathematical and regular than it is. Its exactitude is obvious, but its inexactitude is hidden.
Its wildness lies in wait. In other words, we know what's likely to happen. We know the other things that probably could happen instead.
We have little appreciation for the things that are highly unlikely to happen but could and these are what we call in modern-day terms the tail events. My friend Rick Kane once said that 96% of financial history has occurred within two standard deviations, but everything interesting has happened outside of two standard deviations. That's the wild part. So now let me try in a slightly philosophical sense to reflect to you how I think about risk, how I think you might consider risk through four basic points. Number one, There was a professor at the London Business School who said, risk means more things can happen than will happen.
For most events that lie in the future, there are a number of things that could occur. We don't know which one it will be. That's where the risk comes in. More things can happen than will happen. Number two, as a result of that, the future should be viewed not as a fixed outcome that's destined to happen.
capable of being predicted, but as a range of possibilities and hopefully because you have some insight into their respective likelihoods as a probability distribution. The most likely, the less likely, the unlikely, but not impossible. Number three, it's important to accept that even when you know the probability That doesn't mean you know what's going to happen. This is something that I think many people fail to grasp. I play a lot of backgammon and a lot of my examples on risk and uncertainty.
come from the game of backgammon, which is played with a pair of dice. And when you roll your pair of dice, we know exactly in advance what the probabilities are. Each die has six sides.
There are 36 possible combinations of the six sides. We know how many of them, for example, will add up to seven. And seven is the most likely single outcome.
1, 6, 2, 5, 3, 4, 4, 3, 5, 2, 6, 1. There are six possibilities out of the 30. 36 that will give you a 7. That's the most likely outcome. Six out of 36, that's 16.7 percent probability. Now, what if instead you want to know about a six? Well, with the six, there are five possibilities. Five possibilities out of 36, that's a little less than a seven.
And then when you get down to the number two, there's only one, one, one, one out of 36. And for the number 12, Only one. Six-six. One out of 36. Both of those are about a 3% probability of happening.
So we know exactly what the probability distribution looks like. We know what's the most likely, the other likely possibilities, and the unlikely possibilities. We still don't know what's going to happen.
So knowing the probabilities does not eliminate the uncertainty. I work with a professor at Wharton named Chris Goetze. And the way he put it to me one time. we live in the sample, not the universe.
In other words, the universe statistics, like I just explained for the dice, determine the things that could happen and maybe their possibility, but we live in the sample. We only have one outcome and therein lies the uncertainty. A great way to think about this is on Super Bowl morning in 2016, they had a former football player on. And he said what I thought was one of the smartest things about probability I had ever heard.
This game was Denver versus Carolina. And Carolina was heavily favored. And they asked him who he thought would win.
And he said the following. Carolina wins eight times out of ten. This could be one of the two. Now, this gives you the essence of probability and the essence of risk.
Most people, if they hear... that something's 80% likely to happen, they say, well, then I guess we know what's going to happen. I guess they might as well not play the game. No, 80% likely means that the other team should win one game out of five. So we have to play the game because we have to figure out which game this will be.
And that leads to number four. I take Dimson's statement that risk means more things can happen than will happen, and I turn it over, even though many things can happen, only one will. Thus the expected value, the probability weighted average of the possible outcomes, which is the basis on which people make many decisions, it can be irrelevant.
They take each outcome, they multiply it by the probability, they add them up and they get the expected outcome, and many people will say well we're going to take the course of action that has the highest expected value, but sometimes the expected value value isn't even among the possibilities. Now this sounds highly counterintuitive, but think about this. Let's consider a course of action which has four possible outcomes, two, four, six, and eight.
And let's say that we conclude that each of those four is equally likely to happen. So what we do is we take each one, two, four, six, eight, we multiply it by 25 percent, the possibility of it happening, and we add it together. And in this case, the expected value of 2, 4, 6, 8 is 5. But 5 can't happen.
Remember I said the outcomes can be 2, 4, 6, and 8. So I'm only going through this to show you the possible fallacy of expected value. There's another problem with expected value because even though course of action A can have a higher expected value than course of action B, course of action A may include some possibilities that you just can't live with. Maybe course of action A includes some remote possibility that you lose all your money.
And even though it's highly unlikely, you would. you may say, I just don't want to contemplate that. So you don't take A. You take B instead, which has a slightly lower expected value, but without the risk of ruin. Now, moving on a little bit to talk about the character of risk, I think it's interesting to note that risk is counterintuitive.
They did an experiment in the town of Drachten, Holland. They took away all the traffic lights, traffic signs. and road markings.
What do you think happened to the level of accidents and fatalities? It went down. How could it possibly have gone down when all the road aids were gone?
And the answer is people said, oh oh, there are no more traffic signs, traffic lights, or road markings. I'd better drive more carefully. On the other hand, Jill Fredson is an expert on avalanches, and she said that better gear is created every year, which makes it easier and more feasible to climb, and yet the risk, the number of fatalities and accidents in climbing doesn't go down.
How can... that be? Obviously counterintuitive. People see that better gear is being invented and they say, well since we have better gear we can do riskier things. And the level of accidents and fatalities is maintained even in spite of the arrival of better gear.
So if you think about those two examples you realize that the risk of an activity doesn't just lie in the activity in itself but importantly in how the participants approach it. The degree of risk present in a market or in an investment doesn't come just from the market or the investment, but how people participate in that investment. And if they conclude that the market has become safer, they may say that that frees them to do riskier things. And that's why I believe that risk is low when investors behave prudently and high when they don't.
Just as risk is counterintuitive, I believe that risk is perverse. As I said, the riskiest thing in the world is the belief that there's no risk. A high level of risk consciousness, on the other hand, tends to mitigate risk.
So when people say, oh, that's really risky, if they take a cautious approach, then it becomes safe. As an asset declines in price, most people say, oh, it's risky. Look how it's falling.
But with the lower price it actually becomes less risky. As an asset appreciates, most people say, that's a great asset, look how well it's doing, but the rising price makes it riskier. So again, perverse. And this perversity is one of the main things that render most people incapable of understanding risk. I think it's important to grasp a concept.
Risk is hidden and risk is deceptive. Loss is what happens when risk The potential for loss collides with negative events. You know, Buffett says everything the greatest, and he said that it's only when the tide goes out that we find out who's been swimming naked. It's only in times of testing that investors and their strategies are examined for the risk they really held.
An example of that, I wrote in my book, The Most Important Thing, those of us who live in California, as I did at the time, the time. Our houses might contain a construction flaw, but if all is well, that flaw sits there for year after year and doesn't produce any loss. It's only when the earthquakes occur that the house is tested and the flaws are disclosed and the risk, the potential for loss, turns into actual loss.
So similarly, an investment can be risky, but if it only exists in salutary environments, it may look like a winner for a long time, and it may look safe for a long time. The fact that an investment is susceptible to a risk that occurs extremely rarely, what I call an improbable disaster, what Nassim Nicholas Taleb called the black swan in his excellent book, the infrequency of loss can make it appear— ...that the investment is safer than it really is. And of course that's an entirely risky conclusion. So the infrequency with which risk turns into loss can be deceptive and cause people to underrate the risk involved in an activity. One of the most important things for every investor to learn is that risk is not a function of asset quality.
This too sounds counterintuitive. There's a belief that high quality assets are safe and low quality assets are. risky. I believe quite the opposite.
A high quality asset can be priced so high then it's risky. I came to work in this industry in September of 1969. The banks at that time and I was hired by one of them them engaged in what was called nifty-fifty investing. They invested in what were considered to be the 50 best and fastest growing companies in America, companies so good that nothing bad could ever happen, and there was no price too high for their stocks. And if you bought those great companies the day I got to work in September of 1969, and if you held their stocks tenaciously for the next five years, you lost more than 90 percent of your money because the prices paid were just too high. high and unsustainable, and roughly half of those companies did run into serious fundamental problems.
Their quality alone, or their perceived quality, did not impart to them the safety that people thought it would. And in fact, because people thought they were so safe, they bid them up to prices, which in fact made them risky. On the other hand, a low-quality asset can be cheap enough to be safe. Again, this seems counterintuitive and maybe even... perverse.
When I left the world of equities in 1978, I was asked by Citibank to start their activity in high-yield bonds. And now I was investing in the lowest quality public companies in America and making money steadily and safely. The juxtaposition of these events taught me an important lesson that I want to share with you so that you don't have to learn it firsthand.
My conclusion was it's not what you buy, it's what you pay. And investment success doesn't come from buying good things, but from buying things well. And if you don't know the difference, you have to study up.
There are no assets. that are so good that they can't become overpriced and dangerous. There are very few assets that are so bad that they can't be cheap enough to be attractive as investments. So this is a simple concept, it sounds like to me, but I hope you'll spend a lot of time thinking about its consequences. Now let's talk for a while about the relationship between risk and return.
This is one of the most important of all the topics. When I got to University of Chicago, the Chicago School of Theory with regard to investment had just been developed mostly between... 62 and 64, and I arrived in 67. And there was a graphic that we saw all the time. It shows return on the vertical axis, risk on the horizontal axis, and an upward sloping line to the the right.
We call that a positive correlation. One goes up, the other goes up as well. Now, most people would look at that graphic and say, well, that means two things, that riskier assets have higher returns.
And if you want to make more money, the way to do it is to take more risk. I think that's a terrible formulation. Very simply, if it were true that riskier assets produce higher returns, then they wouldn't be riskier, would they? So that can't be the the right explanation.
What the upward sloping line, the positive correlation, means is that investments that are perceived as being risky have to be perceived as offering higher returns to induce people to make those investments. That makes perfect sense. The only thing is they don't have to deliver. And it's from the possibility that the projected returns will not be delivered that the risk ensues.
When you look at the old graph. The linearity of the relationship between risk and return implies a dependable relationship. And I've always felt that that was misleading. I was never happy when I got out into the real world and had to live with the consequences.
And so I developed a relationship with the risk. My own version of that chart. I took some little bell-shaped probability distributions and I turned them on their side and I superimposed them on the same line.
It's the same underlying line just now with some embellishment. With the old graph, as you move from left to right, the risk increased. and the return increased. But with this new graph, as you move from left to right, the expected return increases just as it did in the old one, but at the same time, the range of possibilities becomes wider and the worst outcomes become worse. That's risk.
This is the way to think about the risk-return relationship. So now let's talk a little more about how risk should be handled. What determines investment success?
And the best way I have to communicate this to you, in my opinion, is like the act of posting. holding one lottery ticket, the outcome, from a bowl full of lottery tickets, the full range of possible outcomes. As Dimson said, we're going to have one outcome.
There could be many outcomes. And the outcome that occurs never amounts, in my opinion, to anything but one ticket pulled from among the many. In my opinion, superior investors have a better sense for the tickets in the bowl, for what proportion of them are winners and what proportion of them are losers, than do most other people. That's what makes them superior.
And thus, they have a better grasp of whether it's worth participating in a any given lottery and how heavily to bet. Now, how should each of us deal with risk? I think that risk is best assessed through subjective judgment.
Since risk cannot be measured, gauging it has to be the province of subject matter experts. And I'm clearly jaundiced on the subject of quantification. I believe imprecise, qualitative, expert opinion about the probability of loss is far more useful.
useful than precise but largely irrelevant numbers concerning past and projected volatility. So what is the essence of risk management? Peter Bernstein, again, my hero, he said, because of the existence of risk, things are going to be different from what we expect from time to time. How well are we prepared to deal when it's different?
This is a great formulation. There's no challenge dealing with the events when they turn out as we expected. The question is, are we prepared for when they don't turn out as expected? According to Bernstein, risk just means things are uncertain. Good things can happen as well as bad things.
But I think the definition of risk should emphasize the bad things. And thus I would say risk is the possibility that from the range of uncertain outcomes and An unfavorable one will be the one that materializes. It can consist of suffering a permanent loss of capital when bad things happen.
It can also consist of missing out on gains when good things happen. These things have to be balanced. Let's say you think that if you buy something today, there's a one-third chance it'll be down in six or 12 months.
What will you do about that risk? Many people will say, well, I just wouldn't buy it. But what do you do about the other two-thirds, the chance that it'll be up in six to 12 months?
How do you balance the two risks? And, you know, in the real world, we can't make decisions in one dimension. We basically have to balance.
I think that risk is something that should be dealt with constantly, continuously, not sporadically. That's why I bridle when I hear this formulation. Is this a risk on market or a risk off market? Remember, risk produces loss when bad things happen, and that's when we need risk control.
But I believe we never know. when bad things will happen and thus when risk control will be needed. I think the right model for thinking about whether we need risk control isn't American football, it's soccer. In American football the team with the ball has the offense on the field.
They have four tries to go ten yards. If they go ten yards they get four more tries to go ten more yards and if they can keep doing it they eventually score. But if the team doesn't go ten yards in four tries, the referee blows the whistle. The ball goes over to the other team and they try to go ten yards in four tries in the opposite direction.
So we have two teams switching between offense and defense, changing personnel when there are stoppages. That has nothing to do with the real world. The right model, as I say, is what the rest of the world calls football.
The same 11 people mostly play the whole game. Nobody tells you when to be on offense or defense, and there are very few stoppages. in which to adjust tactics and personnel. That's the real world.
In investing, one of the key decisions is when to be on offense, when to be on defense, how much to allocate to each of those. But nobody told me. tells you when to do it and nobody stops the game to give you time. I think the best model for investing and risk management is automobile insurance.
We all drive, we all have cars, we all have insurance on our cars, but I don't think any of us get to the end of a year and say I wish I hadn't had insurance because I didn't have an accident. We like having insurance for the safety it give us regardless of whether or not we have an accident in a particular year. I think about the intelligent bearing of risk for profit.
Back in 1981, I was interviewed by one of the first cable networks. And the reporter said to me, how can you invest in high-yield bonds when you know some of them are going to go bankrupt? And for some reason, I was able to come up with the right answer on the spot.
I said, the most conservative companies in America are the life insurance companies. How can they insure people's lives when they know they're all going to die? And I think it's an interesting question.
But the life insurance company is not going to die. insurance company is number one, taking a risk that it's aware of. They're not shocked when somebody dies, that's the way it goes.
Number two, they take a risk they can analyze. And when I was a young man and got my first insurance policy, they sent a doctor to my house to see if I was healthy. Number three, they take a risk that can be diversified.
So no, life insurance companies insure just smokers or just people who live in on the San Andreas Fault or just skydivers, just young people or just old people. They have a mix, a diversified portfolio, and they take a risk that they're well paid to bear. They figure out the probability of what they're going to have to pay you based on actuarial assumptions. They allow some windage for the uncertainty, and then they charge you a premium. We do the same.
We take credit risk that we're aware of. We analyze it. We take a risk that we can diversify. We have large numbers of holdings in every portfolio which respond to different factors, and it's a risk we're well paid to bear. We get what's called a risk premium or a yield premium to take the risk of the fault.
So the bottom line is that I believe risk is kept under control in superior portfolios. That's one of the things that superior investors do. Highly skilled investors assemble portfolios that will produce good returns if things go as expected and resist declines if they don't. This asymmetry is, in my opinion, the critical element, the cornerstone of superior investment. investing.
Assembling a portfolio that incorporates risk control along with the potential for gains is a great accomplishment, but it's often a hidden accomplishment because risk only turns into loss occasionally when the tide goes out. But the prudent investor and hopefully his or her clients knows that risk is being controlled even at times when it doesn't come to the surface. So I think that risk is something to be managed and controlled but not avoided. Risk control is indispensable. Risk avoidance is not an appropriate goal in investing.
Will Rogers said, you've got to go out on a limb sometimes because that's where the fruit is. I think, and my experience tells me from watching others, that risk avoidance equates to return avoidance. Intelligence, intelligent bearing of risk, should be able to enable us to... to make good returns with the risk under control.
So what's the bottom line of all the foregoing? You shouldn't expect to make money without bearing risk. You shouldn't expect to make money just for bearing risk. Risk is best handled on the basis of accurate, subjective... judgments made by experienced expert investors who emphasize risk consciousness.
The great challenge in investing is to limit uncertainty and still maintain substantial potential for gains. And in conclusion, I'll just say that outstanding investors are outstanding for the simple reason that they have a superior sense for the probability distribution that governs future events, the tickets in the bowl, and for whether the potential return compensates for the risks that lurk in the distribution's unattractive left-hand tail. This is what enables them to achieve the asymmetry that characterizes the game.
rises superior investors, participating strongly in the gains when there are gains, and avoiding many of the losses when there are losses. I hope the foregoing discussion will help you be among them.